Speech intelligibility in multitalker situations with vibrotactile spatial cueing

Abstract

Degraded speech intelligibility in multitalker situations can be improved by spatial unmasking, using sound spatialization techniques such as binaural synthesis. However, intelligibility also depends on the ability to focus efficiently on the target. Three experimental sessions explored the benefit of an additional vibrotactile cue, spatialized around the waist, on intelligibility in different spatialized multitalker situations. Results indicate improvements in intelligibility scores and reduced listening effort specifically for an off-center target among masker talkers. Multimodality allows us to better understand the mechanisms of auditory attention and to open up new perspectives for improving speech intelligibility in multitalker situations.